1. Field of the Invention
The present invention relates to a captured image processing system which causes an image output apparatus to output an image captured by a portable terminal apparatus.
2. Description of the Related Art
With the development of Internet technology, chances are increasing to capture images by use of a camera-equipped mobile device such as a mobile phone. Further, chances are also increasing to use a camera-equipped mobile device for image capture of not only landscapes and people, but also explanation diagrams and/or descriptions displayed in various events, and furthermore slides used in an academic conference and the like.
However, a resolution of an image of the entire approximately A4 paper is much lower in a case where the image is captured by a general camera-equipped mobile device is than in a case where the image is scanned by a flathead scanner. For this reason, research is being done in the technique of carrying out a super resolution process with respect to an image captured by a camera-equipped mobile device, so as to obtain a high-definition image.
Note here that the super resolution process is a process in which a high resolution image is prepared from a plurality of offset low resolution images and various methods for carrying out the super resolution process have been suggested such as an ML (Maximum-likelihood) method, a MAP (Maximum A Posterior) method, and the like.
The ML method is a method in which a square error between (i) a pixel value of a low resolution image which pixel value is estimated from a high resolution image vector h and (ii) a pixel value f of an actually observed low resolution image is an evaluation function, so as to assume a high resolution image which minimizes the evaluation function. Assuming that a transformation matrix for estimating the pixel value of the low resolution image from the high resolution image vector h is A, the evaluation function is represented by the following equation (1).
                    [                  Math          .                                          ⁢          1                ]                                                                      Evaluation          ⁢                                          ⁢          function                =                              ∑                          i              =              1                        M                    ⁢                                                                                    f                  i                                -                                                      A                    i                                    ⁢                  h                                                                    2                                              Equation        ⁢                                  ⁢                  (          1          )                    
In the equation (1), M denotes the total number of observation points, fi denotes an image value of the low resolution image at an ith observation point, and Ai denotes a transformation matrix for estimating the pixel value of the low resolution image at the ith observation point from the high resolution image vector h.
The MAP method is a method for estimating a high resolution image which minimizes an evaluation function which is obtained by adding probability information of the high resolution image to the evaluation function used in the ML method.
Such a super resolution process requires capturing a plurality of images while displacing an image capture device at smaller intervals than pixel intervals. This often requires a complicated system which is capable of accurately controlling an image capture apparatus and causes a problem that it is difficult to arrange such a system at low cost.
In contrast, Patent Literature 1 discloses the technique in which image capture with pixel sliding is carried out without the need of control for accurately displacing an image capture device at smaller intervals than pixel intervals, so as to prepare a high resolution image by use of a plurality of low resolution images. An image capture system disclosed in Patent Literature 1 includes state change detection means for detecting a change in state of an image capture apparatus and makes it possible to carry out image capture at a timing at which a motion velocity of the image capture apparatus itself is lower than a given value. According to this, image capture with pixel sliding is carried out by use of a random motion such as a camera shake or motion means such as a spring, so as to prepare, by the super resolution process, a high resolution image from a plurality of low resolution images.